Britannia Builder

Yes, that makes sense. I'm surprised that HRMS doesn't have a secure website - it's pretty much expected these days for any commercial operator (I don't have a security cert for my personal www.britanniabuilder.com website, because of the extra cost involved). Perhaps Hornby could be invited to host a copy of the downloads on their secure website? Regards, John

Thanks Paul. I looked at the Windows antivirus log and the reason it gave for blocking the exe file was that it contained 'HackTool:Win32/RemoteAdmin!MTB'. I guess that this is the software that HMRS use when logging in remotely to do diagnostics. 'HackTool' is presumably the virus checker's view of the thing, rather than part of the name! I hadn't realised that HMRS installed this as part of RailMaster, and frankly I'd rather that they didn't. I wouldn't knowingly let them or anyone else log in remotely to my computer. Regards, John

Pasting that link manually into the Chrome address bar does work - it downloads the zip file. I haven't tried unzipping it. Strangely the size of the zip file is exactly the same as the exe file at 70.9 MB. Regards, John

If I click on the link you've just given, Paul, nothing happens - the screen stays unchanged. Is this my Chrome browser causing the link to be blocked? Regards, John

Prompted by this thread, I checked my RM version number and found that I was still on 1.72.0! I thought RM was meant to update automatically, at least for major updates? I then went through an incredible palaver to do the update - the HMRS link didn't seem to do anything, so I tried the backup MEGA site and eventually after turning off virus checking in Windows Defender and turning off safe browsing in Chrome and overriding security warnings I was able to download the installer and do the update. I can't say that I noticed any difference! Regards, John

Do you have power banks installed? I'm wondering if the inrush of current to the power bank when the loco first starts moving is upsetting things. If so, you could try installing the Tornado profile which charges the power bank before the loco starts moving. Regards, John

All fixed now. I tried reading out all the CVs on the programming track under Railmaster, but there's no easy way of updating them again as RM will only update CVs that have been manually changed. However, having read the CVs into the app and reset CV8 to 8 (which fixed the DCC communication problem as expected), it's quite easy to scroll through the CVs in the app, tap on each one that needs updating, and press exit in the CV value screen without changing anything. It only took a few minutes to reset AFC sounds, complex speed curve etc, then re-run auto calibration. There's obviously still an underlying bug somewhere that needs fixing, but I can't think what combination of events causes the loss of communication. It's only ever affected one of my three TXS locos. Regards, John

Can I put in a vote for a DMU sound profile too - I have a Hornby 3 car Class 101, currently running the Class 31 profile. This sounds OK - it's 60 years since I used to commute to school on Class 101s, so I can't really remember what they sound like - just remember the diesel fumes and the vibration! Regards, John

I can't actually see an 'export profile' button in the app - there's one to export function button settings, but that won't help with CVs. I suppose I could read all the CVs in Railmaster on the programming track and then write them back again after resetting the decoder, but this would be a slow process. Regards, John

Thanks Rob, that sounds perfect - I'll give it a go tomorrow. Regards, John

I've just had this problem again, 3 months after it last happened (see my post from 22 May above). I had done lots of CV updating via the app to set a complex speed curve, then reverted to Railmaster and tinkered with the speed factor to get the scale speed realistic again, then was running the loco quite happily under Railmaster/DCC when it suddenly lost control when running at full speed. I'd been half hoping that this would happen again because I had a theory that CV1, the DCC address, might be corrupted, but I read all the CVs back from the decoder into the app and sadly they all look fine. I've tried power cycling the decoder, and restarting Railmaster. The next step which will hopefully fix it is to reset CV8 to 8, but this will lose all my CV changes including the complex speed curve. I don't suppose that there is any way of backing up the CV settings and restoring them after the reset? Regards, John

Thanks Rob. The APROM version is in CVs 47-49 and the SPIROM version is in CVs 202-203, but I can't see a combined version number. Interestingly the APROM version in my charge-when-stationary Tornado is 0.0.5, the same as my other two TXS decoders which were installed a couple of months ago, so perhaps the new charging feature is not in the APROM? Regards, John

Thanks Steve and Rob. When a new version of a profile is released for this or any other reason, will it be possible to update just the APROM code and SPIROM sounds without overwriting all the CVs and functions that have previously been amended by the user? I imagine that this charge-when-stationary feature is a change to the APROM code rather than a feature of the sound file - will it just download the small APROM file if the SPIROM is unchanged? And is the APROM actually different for all the different sound profiles? Regards, John

How will we know when existing profiles have been updated to include the charging when stationary? I have a Bachmann 3MT Tank which needs it - I'm currently running the Black 5 profile, but I would switch to another similar profile if it got the update earlier. Regards, John

I would say that my Tornado is making between 5.5 and 6.5 chuffs per revolution at low speeds. I guess that is about as accurate as can be expected without sensor hardware to synchronise the chuffs precisely to the revolutions. Each cylinder will make two chuffs per revolution, so 4 chuffs for a 2 cylinder and 6 for a 3 cylinder. I believe that 4 cylinder locos such as a Duchess have the cylinders synchronised in pairs, so 4 (double) chuffs per revolution. Incidentally I had run the auto calibration on my Tornado before setting up the speed curve. I don't know if this would affect the result - ie will auto calibration change the mapping between speed steps and motor speed? Regards, John

I've set up the 3 cylinder speed curve from Rob's post above (5, 6 .... 125, 140) for my Peppercorn A1 with the Tornado profile, and it works very well - 6 chuffs per revolution from the slowest to as fast as I can make them out, and a realistic top speed. I didn't note down the default CV values in the Tornado profile before changing them, but they were nearly twice the above values and the chuffing was way too slow at all but the slowest speed, and the top speed was dangerously high! I'll try my Bachmann BR Standard 3MT Tank next with the 2/4 cylinder values above, and/or LuLuJo's alternatives. Regards, John

That's interesting, and a good enhancement. I downloaded the Tornado profile yesterday and I can confirm that its power bank does now charge when stationary. You say that this is the default - is there a way of turning it off? The only reasons that I could think of for turning it off would be prior to a profile download to speed up the power cycling (which takes a long time on my Tornado, even with the sounds playing, which they won't be during a download), and to allow conventional CV updating on a programming track. But now that the app allows the complex speed curve CVs to be set up, I can't see much reason for doing the latter. Regards, John

I've just downloaded the Tornado profile to my Peppercorn A1, replacing the FS profile - worked first time, and sounds good. Regards, John

I had this problem twice, over a month ago now, and recovered by setting CV8 to 8. This is inconvenient because all CV changes then have to be reapplied. If it does happen again, I shall refresh all the CVs in the app from the decoder and try to see what's changed. I suspect that the DCC address will have been corrupted, since I can't think of any other CV changes which would render the decoder incommunicado. Regards, John

I've also now tried setting F9 as a running mode F28 sound, and this also fails to sound. Regards, John

Yes, I've just refreshed the CVs in the app from the CVs in the decoder for both locos, and CV65 is set to 2 and 3 respectively, so F9 is on in both cases.

2 for F9, or 3 for F9+F8. The app CV editor shows the individual F numbers so there's no need for binary arithmetic.

I tried both random and sequential when trying to get F9 to sound, but it made no difference. I guess most looped sounds are quite short before they repeat. They seem to run for about 10 seconds under AFC, so plenty of gap before the next sound which is typically at an interval of 20-30 seconds. If AFC looped sounds work as they appear to, it might be worth updating the decoder manual to say how they work rather than advising against using them. Regards, John

Just to be clear, I wasn't suggesting that there is a problem with looped sounds under AFC - they seem to work rather well. Regards, John

I use F28, Automatic Function Control, to get sound effects when stationary and when running, and in general it works very well. However, I can't get F9 (Safety Valve on a steam loco) to sound from within AFC. Other sounds such as F7 Coal Shovel and F8 Injector work fine within AFC, and F9 works fine as a stand-alone function. I've tried it in two of my steam locos, in both cases as a stationary sound - I haven't tried it as a running sound. If it's a bug, it's a rather strange one - perhaps someone else could try it out? As another observation on AFC, the manual advises against using looped sounds (by which I assume that it means toggled functions), but when I do use Coal Shovel or Injector in AFC they sound for about 8-10 seconds and then stop, which seems fine. Within the app, such functions can be redefined as timed with a selectable time duration, which is a nice feature with no equivalent in Railmaster. Doing this has no effect on the duration when the function is used within AFC, since the time duration is presumably held in the app rather than in the decoder. Regards, John